AI Agents & Automation

Prompt Refine was a dedicated tool for enhancing prompt engineering workflows, allowing users to methodically improve their Large Language Model (LLM) prompts. It integrated with various AI models, including OpenAI, Anthropic, Together, and Cohere, providing a versatile environment for prompt development. Key functionalities included comprehensive history tracking to analyze and compare different prompt runs, enabling users to refine their approaches based on past results. The platform also supported the creation and reuse of variables within prompts, streamlining the experimentation process. Users could export their experiments to CSV for further analysis, making it a valuable asset for data-driven prompt optimization. However, the tool has since been shut down.

tree-of-thought-llm is the official open-source implementation of the Tree of Thoughts (ToT) framework, designed for deliberate problem-solving with large language models. This repository, published after the NeurIPS 2023 paper, includes the core code, example prompts, and model outputs, enabling researchers and developers to explore and replicate the ToT methodology. It supports various problem-solving tasks like the game of 24, text generation, and crosswords, offering different thought generation and state evaluation methods. Users can easily set up new tasks and customize prompts, making it a flexible tool for advancing research in LLM reasoning and problem-solving.

Vosk-android-demo offers robust offline speech recognition and speaker identification capabilities specifically designed for Android mobile applications. This tool is built upon the powerful Vosk and Kaldi libraries, ensuring high accuracy and performance without requiring an internet connection. Developers can easily integrate these features into their Android projects, with pre-built binaries available in the releases section to streamline the development process. It's an ideal solution for creating mobile applications that require on-device voice command processing, transcription, or user authentication through voice, providing a reliable and efficient way to handle speech data locally.

vstar is an open-source project offering a PyTorch implementation of the research paper "V*: Guided Visual Search as a Core Mechanism in Multimodal LLMs." This tool is designed for researchers and developers working with multimodal large language models, specifically focusing on enhancing visual search capabilities. It includes pre-trained models for both VQA LLM and visual search, along with comprehensive training datasets derived from LAION-CC-SBU, COCO, and GQA. Users can set up a local Gradio demo for interactive use and evaluate models using the V*Bench benchmark. The project also provides detailed instructions for pre-training and instruction tuning of the VQA LLM, making it a valuable resource for advancing research in guided visual search within LLMs.

XVerse is an online demonstration of an AI image generation tool developed by ByteDance. Users can generate images by providing a textual prompt and up to four reference images, enhancing creative control. The application also offers practical features such as auto-captioning for descriptions and face cropping, which can be useful for refining generated images or preparing them for specific uses. Hosted on Hugging Face Spaces, XVerse provides a platform for exploring advanced image synthesis capabilities.

WizardLM 1.0 Uncensored Llama2 13b GGML is an AI chatbot tool designed for generating text responses to user prompts. Users can input any question or request, and the application aims to provide detailed and helpful answers. While the tool's description highlights its text generation capabilities, the current live website indicates a runtime error preventing its operation. This suggests that the model or its associated files are currently inaccessible or improperly configured, leading to a 'Repository Not Found' error. The tool is hosted on Hugging Face Spaces and is intended for AI model experimentation and chatbot development, potentially for educational purposes and research.

AI2C Technologies AG is a Swiss ETH Zurich spin-off specializing in computational thinking. The company develops breakthrough technologies in real-time continual learning (RT/CL) and automatic model recalibration, which are crucial for advanced computational thinking. Their products power 'Computational Thinking' machines designed to work alongside humans, enhancing decision-making across various domains. By integrating computing innovation, scientific principles, advanced mathematics, algorithms, and multidisciplinary knowledge, AI2C's mission is to contribute to the advancement of artificial general intelligence (AGI). The team comprises scientists, engineers, and business innovators with expertise in computational science, artificial intelligence, fluid mechanics, and nanotechnology.

Aigency is a platform dedicated to AI agents, offering a distinct methodology for their development and deployment. It provides a suite of tools and resources designed to facilitate the creation and implementation of AI agents across diverse applications. The platform emphasizes a unique approach that aims to differentiate itself from traditional AI agent development methods, suggesting a focus on innovation and efficiency in the field. While specific features are not detailed, the core offering revolves around empowering users to build and manage AI agents effectively.

Algernon is a small, self-contained web server written in pure Go, designed for web hosting, application development, and content serving. It provides extensive support for scripting languages like Lua and Teal, and integrates with various database backends including Redis, SQLite, PostgreSQL, MariaDB, MySQL, and BoltDB. The server supports modern web technologies such as HTTP/2 and QUIC, and features built-in rendering for Markdown, Pongo2, Amber, Sass (SCSS), GCSS, and JSX. Algernon also includes Ollama for LLM content generation, rate limiting, graceful shutdowns, and a plugin system, all within a single executable. It's versatile, offering live editing/preview with auto-refresh and working across Linux, macOS, and Windows.

The AI SDK is a free and open-source TypeScript toolkit developed by the creators of Next.js, Vercel. It is designed to simplify the development of AI-powered applications and agents, offering a provider-agnostic API that integrates with popular UI frameworks such as Next.js, React, Svelte, Vue, and Angular, as well as runtimes like Node.js. The SDK supports interaction with major model providers including OpenAI, Anthropic, and Google, often leveraging the Vercel AI Gateway for seamless access. Developers can generate text, structured data, and build complex agents with integrated tools. It also includes a UI module with framework-agnostic hooks for creating chatbots and generative user interfaces.

Qualcomm® AI Hub Models offers a comprehensive collection of machine learning models specifically optimized for deployment on Qualcomm® devices. This includes a wide array of models across categories like Computer Vision (Image Classification, Image Editing, Super Resolution, Semantic Segmentation, Video Classification, Video Generation, Video Object Tracking, Object Detection, Pose Estimation, Gaze Estimation, Depth Estimation, Driver Assistance, Robotics) and Multimodal tasks. The models are designed for high performance, low latency, and efficient memory usage on various Qualcomm® chipsets and devices. Users can install the Python package, configure AI Hub Workbench access for compilation and profiling, and export/run models on physical devices or utilize end-to-end demos. The platform supports multiple on-device runtimes and hardware targets, making it a versatile resource for developers working with Qualcomm® AI hardware.

The Hebrew LLM Leaderboard is a Hugging Face Space designed for evaluating and comparing the performance of Hebrew large language models. Users can explore a comprehensive leaderboard that is both searchable and filterable, allowing for detailed analysis of benchmark results. The platform offers customization options, enabling users to select which columns to display and to filter models by type, size, and precision. This tool is invaluable for researchers, developers, and students interested in the advancements and capabilities of Hebrew LLMs, providing a clear overview of model performance on diverse tasks. It is freely available and serves as a critical resource for language research and educational purposes within the AI community.

ChatGPT Adapter is a versatile service designed to unify access to various AI chat interfaces under a standard OpenAI API. It integrates popular AI models such as OpenAI API, Coze, DeepSeek, Cursor, Windsurf, Qodo, Blackbox, You, Grok, and Bing Drawing. This adapter allows developers to leverage a wide range of AI capabilities through a single, familiar interface, simplifying development and integration. Key features include support for high-speed streaming output and seamless multi-turn conversations, ensuring compatibility with existing OpenAI API workflows. The project also provides resources for understanding reverse engineering techniques for AI interfaces, catering to users interested in the underlying mechanisms.

ChatGPT-API-Faucet is an open-source project designed to support AI ecosystem developers by providing free ChatGPT API tokens. Inspired by cryptocurrency faucets, this platform allows users to claim one token every 24 hours, which can be used for developing and testing AI products. The project's frontend is built using Next.js and React, making it a suitable resource for developers looking to experiment with AI APIs without immediate cost. It offers a practical solution for those needing small amounts of API credit for initial development, prototyping, or educational purposes, fostering innovation within the AI community.

Sdcb Chats is a robust and adaptable frontend and AI gateway designed for large language models, supporting more than 22 mainstream AI model providers. It offers a unified management solution for various model interfaces, simplifying deployment with a single Docker command. Key features include a code interpreter with integrated tools like a browser and Excel, an API gateway compatible with Chat Completions/Messages, and support for multimodal inputs and image generation. The platform also boasts enterprise-grade security with user permission management, account balance control, rate limiting, audit logs, and support for Keycloak SSO and SMS verification login. It provides comprehensive observability with full-link request tracing for quick issue identification.

KAN-TTS is a comprehensive speech-synthesis training framework designed to empower users to develop and customize their own text-to-speech (TTS) models from the ground up. The framework currently supports popular models such as sam-bert and hifi-GAN, with plans to integrate more in the future. It offers extensive language support, including Mandarin, English, British English, Shanghainese, Sichuanese, Cantonese, Italian, Spanish, Russian, and Korean, making it versatile for a global audience. KAN-TTS provides a training tutorial through its wiki page and offers a demo on ModelScope for users to experience its capabilities. The project is open-source, hosted on GitHub, and encourages community contributions.

Kernel Labs is a startup studio located in Seattle, Washington, with a strong focus on machine learning, computer vision, and security. The company's mission is to build innovative ventures that aim to create and disrupt significant markets. They actively partner with CEOs, providing deep expertise and support to nurture new ideas from concept to fruition. Their team comprises industry veterans with a proven track record in technology and business development, including a CEO passionate about disruptive innovation and a VP of Engineering with a Ph.D. in Computer Science. Kernel Labs is dedicated to developing technology strategies and ensuring excellence in product development for their portfolio companies.

KVCache-Factory is a unified framework designed for KV Cache compression methods specifically for auto-regressive models. It offers support for multi-GPU inference, making it suitable for large language models such as Llama-3-70B-Instruct. The framework integrates various compression techniques including PyramidKV, SnapKV, H2O, and StreamingLLM, and is compatible with Flash Attention v2 and Sdpa Attention. It provides tools for performance visualization and supports inference on benchmarks like LongBench and Needle in a Haystack. KVCache-Factory is an open-source project, making it accessible for developers and researchers working on optimizing LLM inference.

LLM-Dojo is a lightweight, open-source framework designed for post-training large language models (LLMs). It offers comprehensive support for various training methodologies, including Supervised Fine-Tuning (SFT), Reinforcement Learning from Human Feedback with Value Regularization (RLVR), On-Policy Knowledge Distillation (On-Policy KD), and Guide Knowledge Distillation (Guide KD). The platform also facilitates mixed training approaches, enabling single-round or multi-round Guide distillation, multi-teacher distillation, and reward mixed training. A key feature is its automated data shunting capabilities. Built on a refactored OpenRLHF core, LLM-Dojo streamlines the framework by retaining only the essential RLVR components and integrating advanced KD and Guide-KD techniques, making it suitable for rapid fine-tuning experiments with features like Deepspeed support, LoRA/QLoRA, and automatic chat template adaptation.

free-gpt3.5-2api is an open-source project designed to offer free API access to GPT-3.5, enabling developers to integrate powerful language models into their applications. It supports various authentication methods, including免登录chat2api,账号chat2api, and ACCESS_TOKEN, providing flexibility for different use cases. The tool can be easily deployed using Docker, Docker Compose, Vercel, or Koyeb, making it accessible for a wide range of development environments. It also includes features to prevent API abuse, ensuring secure and controlled access. The project offers model mapping for various GPT-3.5 turbo versions to a common render-sha model, and also supports gpt-4o.

Hugging Multi-Agent is a comprehensive tutorial designed for developers interested in understanding and implementing multi-agent systems, particularly those based on the MetaGPT framework. It offers a practical learning path, guiding users from foundational agent concepts to the development of complex multi-agent applications. The tutorial is ideal for engineers aiming for career advancement in large language model and agent development, focusing on hands-on coding and personalized agent capabilities. It requires Python programming skills, including some asynchronous programming knowledge, and the ability to read and understand project source code. The resource covers agent structure, multi-agent frameworks, and practical development steps, including creating simple and multi-functional agents, as well as managing agents.

EIDON AI offers a comprehensive data infrastructure layer for robotics, focusing on collecting and processing human demonstration data for AI robot manipulation. The platform includes the Eidon Tracker, a 7-IMU wearable for full upper-body arm kinematics, and the Eidon Glove, which provides 16-DOF finger tracking. Data collection is facilitated by the Eidon App, available on iOS and Android, which syncs natively with the hardware to capture synchronized egocentric video and sensor data. This app also supports video-only collection and handles operator payments. Collected data flows into Eidon Sym, a simulation environment and data pipeline that uses VLM-powered quality control to filter, auto-tag objects, and output simulation-compatible formats for model training.

PyOD is a comprehensive Python library for anomaly detection, established in 2017 and widely used in both academic research and commercial products. It supports over 60 detectors across tabular, time series, graph, text, and image data, all accessible through a unified API. Version 3 introduces ADEngine for intelligent orchestration and an agentic workflow via the 'od-expert' skill for AI agents, allowing natural language interaction for anomaly detection investigations. The library maintains backward compatibility with its classic fit/predict API and is built on SUOD for fast parallel training and Numba JIT for per-model speedups. It is recognized for its impact in space and science, enterprise deployments, and educational courses.

pytorch-seq2seq offers comprehensive tutorials for understanding and implementing sequence-to-sequence (seq2seq) models using the PyTorch deep learning framework and TorchText library. The repository focuses on practical application, guiding users through the process of training models for neural machine translation, specifically from German to English. It covers foundational seq2seq concepts, including encoder-decoder models with LSTMs and GRUs, and delves into advanced topics like attention mechanisms to alleviate information compression problems. The tutorials are structured to build knowledge progressively, starting with basic workflows and moving to more sophisticated architectures. It also provides necessary setup instructions, including dependency installation and spaCy model downloads, making it a valuable resource for those looking to implement and experiment with seq2seq models.